Outcomes of Pediatric Macular Hole

BPEI Biostatistics Center

Author

Lomas Persad

Published

April 11, 2025

Abstract

Hypothesis: Macular hole surgery for pediatric macular hole has favorable visual outcomes and a low rate of complications. However, patients managed with observation may also have favorable visual outcomes.

Rationale: Pediatric macular holes are less common than adult macular holes and usually are discovered in the setting of trauma. Though some pediatric macular holes may close with observation alone, others may require surgery. As this is a rare condition, clinical natural course and history studies are limited. The purpose of this study is to explore the clinical outcomes associated with pediatric macular hole in a large dataset.

Data extraction

graph TD
    A[Pediatric macular hole dx between Jan 2013 - Oct 2024 <br> 3098 eyes, 2763 patients]
    A --> B[Exclusions due to concomitant dx: <br> 348 eyes, 258 patients<br>Loss to F/U:  1217 eyes, 1101 patients<br> Unspecified eye: 145 eyes, 145 patients<br> Missing demographic data : 704 eyes, 636 patients ]
    
    B --> C[Remaining: 684 eyes, 623 patients]
    C --> D[Surgery group: 129 eyes, 129 patients]
    C --> E[Observation group: 555 eyes, 494 patients]
    
    D --> F[Surgical Complications: 24 patients]
    E --> G[Surgical Complications: 22 patients]

Summary

Table 1: Demographic data of surgical and non-surgical patients

Variable

Overall
N = 6231

Observation
N = 4941

Surgery
N = 1291

p-value2

Sex

0.3

Female

219 (35%)

179 (36%)

40 (31%)

Male

404 (65%)

315 (64%)

89 (69%)

Both eyes included

61 (9.8%)

58 (12%)

3 (2.3%)

0.001

Age at first macular hole diagnosis

14.00 (11.00, 16.00)

14.00 (11.00, 16.00)

14.00 (11.00, 15.00)

0.8

Age at macular hole diagnosis for fellow eye

13.00 (9.00, 16.00)

13.00 (9.00, 16.00)

13.00 (11.00, 16.00)

0.7

Unknown

562

436

126

Race

0.5

Asian

20 (3.2%)

18 (3.6%)

2 (1.6%)

Black Or African American

99 (16%)

74 (15%)

25 (19%)

Other

108 (17%)

87 (18%)

21 (16%)

White

396 (64%)

315 (64%)

81 (63%)

Ethnicity

0.3

Hispanic Or Latino

105 (17%)

79 (16%)

26 (20%)

Not Hispanic Or Latino

518 (83%)

415 (84%)

103 (80%)

Region

>0.9

Midwest

143 (23%)

111 (22%)

32 (25%)

Northeast

122 (20%)

98 (20%)

24 (19%)

South

267 (43%)

214 (43%)

53 (41%)

West

91 (15%)

71 (14%)

20 (16%)

Follow up (months)

35.45 (15.44, 64.16)

38.58 (16.62, 67.64)

24.38 (13.50, 47.14)

<0.001

Note: Three patients are count in both groups as they had one eye in each group

1n (%); Median (Q1, Q3)

2Pearson's Chi-squared test; Wilcoxon rank sum test; Fisher's exact test

Observation group

Table 2: Surgical complication for observation group

Variable

No surgical complication
N = 4751

Complex RD
N = 111

Photocoagulation
N = 81

Scleral buckling
N = 11

Vitrectomy
N = 21

p-value2

Sex

0.8

Female

170 (36%)

4 (36%)

3 (38%)

1 (100%)

1 (50%)

Male

305 (64%)

7 (64%)

5 (63%)

0 (0%)

1 (50%)

Both eyes included

54 (11%)

0 (0%)

3 (38%)

1 (100%)

0 (0%)

0.025

Bilateral surgery

0 (NA%)

0 (0%)

2 (25%)

1 (100%)

0 (0%)

0.056

Unknown

475

0

0

0

0

Age at first macular hole diagnosis

14.00 (11.00, 16.00)

14.00 (12.00, 16.00)

11.00 (10.00, 14.50)

17.00 (17.00, 17.00)

14.00 (11.00, 17.00)

0.4

Age at macular hole diagnosis for fellow eye

13.00 (9.00, 16.00)

NA (NA, NA)

13.00 (10.00, 17.00)

17.00 (17.00, 17.00)

NA (NA, NA)

0.3

Unknown

421

11

5

0

2

Time to complication (days)

NA (NA, NA)

34.00 (16.00, 893.00)

30.00 (10.00, 1,463.00)

303.00 (303.00, 303.00)

217.00 (22.00, 412.00)

>0.9

Unknown

475

0

1

0

0

Time to complication, fellow eye(days)

NA (NA, NA)

NA (NA, NA)

702.00 (623.00, 1,161.00)

681.00 (681.00, 681.00)

NA (NA, NA)

0.7

Unknown

475

11

5

0

2

Race

0.038

Asian

16 (3.4%)

0 (0%)

2 (25%)

0 (0%)

0 (0%)

Black Or African American

70 (15%)

3 (27%)

1 (13%)

0 (0%)

0 (0%)

Other

82 (17%)

5 (45%)

2 (25%)

0 (0%)

0 (0%)

White

307 (65%)

3 (27%)

3 (38%)

1 (100%)

2 (100%)

Ethnicity

0.2

Hispanic Or Latino

74 (16%)

3 (27%)

2 (25%)

1 (100%)

0 (0%)

Not Hispanic Or Latino

401 (84%)

8 (73%)

6 (75%)

0 (0%)

2 (100%)

Region

0.9

Midwest

108 (23%)

2 (18%)

1 (13%)

0 (0%)

0 (0%)

Northeast

97 (20%)

1 (9.1%)

2 (25%)

0 (0%)

0 (0%)

South

202 (43%)

5 (45%)

5 (63%)

1 (100%)

2 (100%)

West

68 (14%)

3 (27%)

0 (0%)

0 (0%)

0 (0%)

Follow up (months)

37.42 (16.20, 66.69)

56.31 (35.97, 87.75)

72.62 (37.17, 88.26)

58.05 (58.05, 58.05)

72.60 (34.79, 110.41)

0.2

1n (%); Median (Q1, Q3)

2Fisher's exact test; Kruskal-Wallis rank sum test

Surgical group

Table 3: Surgical complication for Surgical group

Variable

No surgical complication
N = 1051

Cataract Surgery
N = 61

Macular Hole Surgery
N = 141

Retinal Detachment Surgery
N = 41

p-value2

Sex

0.5

Female

32 (30%)

2 (33%)

6 (43%)

0 (0%)

Male

73 (70%)

4 (67%)

8 (57%)

4 (100%)

Age at first macular hole diagnosis

14.00 (11.00, 15.00)

16.00 (15.00, 17.00)

10.50 (7.00, 13.00)

12.50 (11.50, 13.50)

0.005

Time to complication (days)

NA (NA, NA)

532.00 (268.00, 725.00)

69.00 (21.00, 147.00)

38.00 (20.00, 73.00)

0.001

Unknown

105

0

0

0

Race

0.4

Asian

1 (1.0%)

0 (0%)

0 (0%)

1 (25%)

Black Or African American

20 (19%)

1 (17%)

4 (29%)

0 (0%)

Other

18 (17%)

1 (17%)

2 (14%)

0 (0%)

White

66 (63%)

4 (67%)

8 (57%)

3 (75%)

Ethnicity

0.2

Hispanic Or Latino

19 (18%)

2 (33%)

3 (21%)

2 (50%)

Not Hispanic Or Latino

86 (82%)

4 (67%)

11 (79%)

2 (50%)

Region

0.9

Midwest

25 (24%)

2 (33%)

4 (29%)

1 (25%)

Northeast

20 (19%)

0 (0%)

2 (14%)

2 (50%)

South

44 (42%)

3 (50%)

5 (36%)

1 (25%)

West

16 (15%)

1 (17%)

3 (21%)

0 (0%)

Follow up (months)

23.46 (13.34, 46.12)

27.45 (19.05, 49.74)

43.64 (11.27, 74.74)

16.06 (12.12, 23.55)

0.5

Note: No one from the surgical group had a Endophthalmitis diagnosis.
One patient had bilateral macular hole diagnosis, both eyes had surgery (133 days apart) but no complication

1n (%); Median (Q1, Q3)

2Fisher's exact test; Kruskal-Wallis rank sum test

VA comparison

These results suggest that while surgical patients started with worse VA, both groups experienced comparable gains in vision over time.

Key findings

Category

Key Findings

Baseline Difference

- Surgical patients had 48% worse VA at baseline compared to the observed group (Exp(β) = 1.48, p < 0.001).

Time Effects

- Both groups showed significant improvement at all follow-up timepoints (p < 0.01)

- Peak improvement occurred at 6 months (25% improvement, Exp(β) = 0.75, p < 0.001)

- No significant interaction between group and timepoint (LRT p = 0.43), indicating similar improvement patterns for both groups over time

Figure 1: Intra group
Figure 2: Inter group

GEE model

Variable

Multiplicative Effect

p-value

Variable

Exp(β)

95% CI (Lower)

95% CI (Upper)

p-value

(Intercept)

0.7944374

0.7350682

0.8586016

<0.001

group_typeSurgical

1.4773713

1.2746061

1.7123925

<0.001

timepoint3MO

0.7976433

0.7383788

0.8616646

<0.001

timepoint6MO

0.7523988

0.6874750

0.8234538

<0.001

timepoint1Y

0.7860029

0.7221201

0.8555372

<0.001

timepoint2Y

0.8157669

0.7274438

0.9148138

<0.001

GEE Forest plot

GEE Forest plot

Time to max BCVA

Figure 3: Time to max BCVA
Figure 4: Change in VA from T1 to max BCVA

Logistic Mixed Effects Model

Table 4: Multivariate data for Logistic mixed effects model

Variable

Overall
N = 5931

Observation
N = 4781

Surgery
N = 1151

p-value2

Age at macular hole diagnosis

14 (11, 16)

14 (11, 16)

14 (11, 16)

0.9

Sex

0.2

Female

214 (36%)

178 (37%)

36 (31%)

Male

379 (64%)

300 (63%)

79 (69%)

Race

0.4

Black Or African American

100 (17%)

76 (16%)

24 (21%)

Other

121 (20%)

100 (21%)

21 (18%)

White

372 (63%)

302 (63%)

70 (61%)

Ethnicity

0.5

Hispanic Or Latino

95 (16%)

74 (15%)

21 (18%)

Not Hispanic Or Latino

498 (84%)

404 (85%)

94 (82%)

Region

0.5

Midwest

135 (23%)

103 (22%)

32 (28%)

Northeast

126 (21%)

103 (22%)

23 (20%)

South

246 (41%)

200 (42%)

46 (40%)

West

86 (15%)

72 (15%)

14 (12%)

VA (logMAR)

0.54 (0.18, 1.00)

0.48 (0.10, 1.00)

1.00 (0.54, 1.18)

<0.001

VA groups

<0.001

Normal

204 (34%)

195 (41%)

9 (7.8%)

Mild

128 (22%)

100 (21%)

28 (24%)

Moderate

138 (23%)

93 (19%)

45 (39%)

Severe

123 (21%)

90 (19%)

33 (29%)

1Median (Q1, Q3); n (%)

2Wilcoxon rank sum test; Pearson's Chi-squared test

Vision impairment (as classified by the vision_group variable) is a strong predictor of surgery, with those in the “Mild,” “Moderate,” and “Severe” vision groups having much higher odds of surgery compared to those with “Normal” vision. The severity of vision impairment increases the odds of surgery.

Other variables such as age, sex, and ethnicity do not significantly predict whether a patient has surgery.

Overall, these results provide evidence that patients with worse vision are more likely to undergo surgery, supporting the notion that visual acuity at baseline plays a crucial role in surgical decisions.

Uni vs Multivariate analysis

Univariate Models

Multivariate Model

Univariate Analysis

OR

95% CI

p-value

OR

95% CI

p-value

ageatdx

1.01

0.76, 1.36

0.926

1.02

0.95, 1.10

0.554

sex

Female

—

—

—

—

Male

1.30

0.16, 10.8

0.807

1.17

0.70, 1.93

0.552

race

Black Or African American

—

—

—

—

Other

0.60

0.02, 14.8

0.753

0.57

0.25, 1.33

0.194

White

0.70

0.06, 8.55

0.781

0.65

0.34, 1.25

0.193

ethnicity

Hispanic Or Latino

—

—

—

—

Not Hispanic Or Latino

0.81

0.06, 10.8

0.874

0.58

0.29, 1.18

0.133

region

Midwest

—

—

—

—

Northeast

0.66

0.04, 12.4

0.784

0.77

0.38, 1.57

0.477

South

0.71

0.06, 8.09

0.786

0.64

0.34, 1.21

0.169

West

0.60

0.02, 17.2

0.767

0.54

0.23, 1.25

0.149

vision_group

Normal

—

—

—

—

Mild

6.52

2.76, 15.4

<0.001

7.33

3.01, 17.8

<0.001

Moderate

11.7

4.76, 29.0

<0.001

12.2

4.77, 31.2

<0.001

Severe

8.69

3.62, 20.9

<0.001

8.68

3.53, 21.3

<0.001

Abbreviations: CI = Confidence Interval, OR = Odds Ratio